INFLUENCE OF ENVIRONMENT 255 



larvae may develop. Sometimes a lily leaf will carry a film of water on 

 its surface in which mosquito larvae can develop safely screened from 

 hungry fish. Naturally the removal of such screens stops mosquito de- 

 velopment by allowing the fish to find the wrigglers. 



Dr. Smith was greatly impressed with possibilities of the top minnow 

 (Gambusia affinis) as a destroyer of mosquito larvae. In his report for 

 1904 he has the following to say about this species : 



"This is a very active species, being constantly in motion, a top 

 feeder, and penetrates to the shallowest edges of the waters it inhabits. 

 Being very small and slender, it easily glides over masses of aquatic 

 plants and penetrates among them as no other fish does. It has been 

 tested in small ponds and tubs during two summers and no mosquito 

 larvae make their appearance when it is present, while Anopheles will 

 appear in similar places where there are goldfish and other fish. Two 

 top minnows will keep a tub clear of mosquito larvae. They breed rap- 

 idly, bringing forth young several times in a season. When extruded, 

 they are many times larger than the newly hatched goldfish and, hav- 

 ing no umbilical sac, they begin to prey at once. 



"Among the fishes of New Jersey there is none that quite fills the 

 place of this Gambusia for choked-up streams and swampy areas. It 

 lives in fresh water as readily as it does in brackish, and does well in 

 captivity or restricted areas generally. Its ability to get along in the 

 merest film of water over a lily leaf, where Anopheles larvae find safety 

 from other fish, gives it a field of usefulness whose importance cannot be 

 overestimated. 



"Mr. Seal writes on this point : 'The seacoast of New Jersey is, in 

 my opinion, amply protected so far as fishes are concerned. I do not 

 think you can improve on Fundulus, Cyprinodon, and Lucania for the 

 marshes of the coast. And as for the shore waters of the Delaware and 

 its open tributaries, Fundulus will do the work. But there are thousands 

 of 'branches' which connect with the creeks by sluices, and which are 

 sluggish and dense with plants. This, therefore, is a separate problem 

 from that of the coastal waters proper. 



"It is my belief that it would be wise to place Gambusia in the head 

 water 'branches' of the cedar swamp-fed streams, such as the Rancocas, 

 Big and Little Egg Harbors, and others, and from these small spring- 

 fed tributaries, many of which never freeze over, let them spread down 

 the streams. I am satisfied that if we go about it right, there will be 

 no trouble in introducing them to such waters ; and in thinking the mat- 

 ter over in connection with such places as I know of as being suitable, 

 I conclude that spring would be much the best time for the experiment. 



